The present invention relates to a computer system comprising a disk unit and, more particularly, to control of logical volume duplication.
In public and private mission-critical systems constituting social infrastructure, the amount of data has recently been significantly increasing. These systems, of course, have a demand for high availability. For this reason, there is often used a storage system based on a disk array technology typified by RAID (Redundant Arrays of Inexpensive Disks) as shown in Japanese Patent Laid-Open Publication No. 2001-256003. In many cases, data for business affairs to be conducted by such a system is placed on a plurality of volumes because of its amount, and a group of volumes used for each of kinds of business affairs are generally handled as one group. Because of this, there have emerged management techniques for efficiently managing volumes, the number of which is increasing with an increase in data amount. The techniques include one of Japanese Patent Laid-Open Publication No. 2005-157710 which realizes power savings by powering off some of disks.
In each of the prior-art techniques, a group of data volumes for each of kinds of business affairs to be conducted is placed with emphasis on performance when it is to be placed in a storage system. For example, separate volumes which belong to a group are placed in different storage systems or array groups, thereby distributing the load on each disk controller and achieving an increase in I/O speed. If the technique shown in Japanese Patent Laid-Open Publication No. 2005-157710 is applied to this storage system, the volumes within the group are distributed over the plurality of storage systems. Accordingly, the volumes cannot be powered off, and the effect of saving power cannot be obtained. In particular, if each volume is multiply duplicated to prevent any data loss caused by a failure in a storage, the range in which power cannot be turned off may extend with an increase in the number of duplicates.
Assume that all of the volumes of the duplicated group are powered off. In accessing the volumes again, there is a waiting time before the all of target volumes are powered on. The waiting time becomes long depending on the placement status of the volumes.